Do different types of dynamic extrapolation rely on the same mechanism?

Observers can estimate the position of occluded moving objects, and this ability could be mediated by the oculomotor control system. However, other forms of motion cannot be tracked with pursuit eye movements. In Experiment 1, 2 kinds of motion extrapolation tasks were interleaved. In the Position extrapolation trials, participants observed a moving target that became occluded. They attempted to press a button when it arrived at the end of the occluder. In novel Accumulation extrapolation trials, they observed a matrix of Gabors with new elements appearing over time. The participants pressed when Gabors would have filled the entire matrix, had they continued accumulating at the same rate. In both tasks completion time estimates were related to presentation parameters in the same way. Moreover, there were robust intertask correlations: Participants who responded earlier in the position task also responded earlier in the accumulation task. Experiment 2 replicated these results with a third form of extrapolation. Although performance was not identical in all tasks, there were enough similarities to suggest that overlapping systems guide all forms of extrapolation. We propose that a common rate control mechanism guides extrapolation, like the velocity store in oculomotor control, but with a broader function than previously envisaged.